Flutter aplication which detect liveness and capture the selfie send to webservice (POST) -- 2
Budget: $10 – $30 USD
I need rebuild from scratch flutter / dart app solution take cam pic from mobile using phone camera and do antispoofing actions nerds detect if user try hijacking/hacking or use a static photo / mask or someting else to bypass it.
If live prove is valid needs send the picture to a rest server using POST method a static shot to a webservice and i will procede with server side validation of this picture (enroll process) the webservice returns a token (randon number) will be displayed to user will use on app.
The app will have 2 tabs one with de face capture, validation and token display and another tab with webview with a website app. More details on video i have apk of similar app but no have the source code the drive is corrupted with no backup.
One method todo it:
Single Face detection randomic asks to user to take serious, blink eyes or smile using random sequence beetween this 3 face types using random steps 2-5 (minimal 2 max 5) requests beetween this actions or using another methods to avoid face spoffing.
Example 1: Take Serious, Smile, Blink
Example 2: Smile, Take Serious, Blink
Example 3: Take Serious,Smile
My research i have some libs and packages todo this job more easy and improve the security.
https://developer.huawei.com/consumer/en/doc/development/HMS-Plugin-Guides/liveness-detection-0000001058389193
https://pub.dev/packages/selfie_ocr_mtpl
https://pub.dev/packages/baidu_face_plugin/example
If live prove is valid needs send the picture to a rest server using POST method a static shot to a webservice and i will procede with server side validation of this picture (enroll process) the webservice returns a token (randon number) will be displayed to user will use on app.
The app will have 2 tabs one with de face capture, validation and token display and another tab with webview with a website app. More details on video i have apk of similar app but no have the source code the drive is corrupted with no backup.
One method todo it:
Single Face detection randomic asks to user to take serious, blink eyes or smile using random sequence beetween this 3 face types using random steps 2-5 (minimal 2 max 5) requests beetween this actions or using another methods to avoid face spoffing.
Example 1: Take Serious, Smile, Blink
Example 2: Smile, Take Serious, Blink
Example 3: Take Serious,Smile
My research i have some libs and packages todo this job more easy and improve the security.
https://developer.huawei.com/consumer/en/doc/development/HMS-Plugin-Guides/liveness-detection-0000001058389193
https://pub.dev/packages/selfie_ocr_mtpl
https://pub.dev/packages/baidu_face_plugin/example